Electronic device, and method of evaluating task performed on document

Abstract

An electronic device according to various embodiments of the present document may comprise a memory and at least one processor operatively connected to the memory. The memory may store instructions that may be executed by the at least one processor and, when executed, cause the electronic device to: obtain an original document and a result document generated by performing a task on the original document; determine evaluation items for evaluating the task performed to generate the result document; generate a prompt that includes the original document, the result document, and at least some of the determined evaluation items and is used to request an AI model to evaluate the task; transfer the generated prompt to the AI model; and obtain a first evaluation result for the performed task from the AI model. Various other embodiments are possible.

Description

How to evaluate tasks performed on electronic devices and documents

[0001] This article is about electronic devices, and for example, how electronic devices evaluate tasks performed on a particular document.

[0002] Generative AI models (GAIs) are AI models that learn from various data and generate new information and sentences. Generative AI models (hereinafter referred to as AI models) are used in natural language processing, enabling them to understand the context of given information and perform various language tasks. For example, a user can generate prompts and input them into the AI ​​to perform a desired task. These prompts can then guide the AI ​​model to perform the desired task.

[0003] Electronic devices such as smartphones, tablets, and laptops can provide a variety of services using various applications. An example of a service that can be provided by an electronic device is a service that performs various tasks, such as summarizing and formatting, on an original document and then generates a resulting document. Furthermore, it may be necessary to evaluate the success of the resulting document generated by the electronic device performing a specific task.

[0004] Traditionally, manual evaluation of resulting documents has been common. This approach can lead to human error, and as the number of documents to be evaluated increases, the evaluation time and resources required can increase. Furthermore, document evaluation software can only provide evaluations of extracted summaries. For generated summaries containing other words and information, the evaluation results may be inaccurate due to the small number of matching words.

[0005] An electronic device according to the present disclosure (or specification, invention) may include a memory and at least one processor operatively connected to the memory.

[0006] According to one embodiment, the memory may store instructions that are executable by at least one processor and, when executed, cause the electronic device to obtain an original document and a result document generated by performing a task on the original document, determine evaluation items for evaluating the task performed to generate the result document, generate a prompt for requesting an AI model to evaluate the task, including at least some of the original document, the result document, and the determined evaluation items, transmit the generated prompt to the AI ​​model, and obtain a first evaluation result for the performed task from the AI ​​model.

[0007] A method performed by an electronic device according to various embodiments of the present document may include an operation of obtaining an original document and a result document generated by performing a task on the original document, an operation of determining evaluation items for evaluating the task performed to generate the result document, an operation of generating a prompt for requesting an AI model to evaluate the task, the prompt including at least some of the original document, the result document, and the determined evaluation items, an operation of transmitting the generated prompt to the AI ​​model, and an operation of obtaining a first evaluation result for the performed task from the AI ​​model.

[0008] A computer-readable non-transitory recording medium according to various embodiments of the present document may store instructions for performing an operation of obtaining an original document and a result document generated by performing a task on the original document, an operation of determining evaluation items for evaluating the task performed to generate the result document, an operation of generating a prompt for requesting an AI model to evaluate the task, the prompt including at least some of the original document, the result document, and the determined evaluation items, an operation of transmitting the generated prompt to the AI ​​model, and an operation of obtaining a first evaluation result for the performed task from the AI ​​model.

[0009] According to various embodiments of the present document, an electronic device and a method for evaluating a task performed on a document can be provided, which can increase the accuracy of the evaluation and reduce the resources used for the evaluation by evaluating the resulting document generated by performing a task on an original document using an AI model.

[0010] FIG. 1 is a block diagram of an electronic device within a network environment according to various embodiments.

[0011] FIG. 2 is a block diagram illustrating an integrated intelligence system according to one embodiment.

[0012] FIG. 3 is a diagram showing a form in which relationship information between concepts and actions is stored in a database according to one embodiment.

[0013] FIG. 4 is a block diagram of a generative artificial intelligence system according to one embodiment.

[0014] FIG. 5 illustrates an electronic device and an AI model according to one embodiment.

[0015] Figure 6 is a block diagram of an electronic device according to one embodiment.

[0016] Figure 7 is a block diagram of a document evaluation system according to one embodiment.

[0017] FIG. 8 illustrates a method for evaluating the performance results of a task from a source document and a result document according to one embodiment.

[0018] Figure 9 illustrates an evaluation item database according to one embodiment.

[0019] Figure 10 illustrates information entered into a prompt for document evaluation according to one embodiment.

[0020] Figure 11 illustrates a prompt generated for document evaluation according to one embodiment.

[0021] Figure 12 illustrates a response of an AI model according to one embodiment.

[0022] Below, various embodiments are described in detail with reference to the drawings so that those skilled in the art can easily implement the present disclosure. In the description of the drawings, the same or similar reference numerals may be used to refer to identical or similar components. Furthermore, in the drawings and related descriptions, descriptions of well-known functions and configurations may be omitted for clarity and conciseness.

[0023] FIG. 1 is a block diagram of an electronic device (101) within a network environment (100) according to various embodiments.

[0024] Referring to FIG. 1, in a network environment (100), an electronic device (101) may communicate with an electronic device (102) via a first network (198) (e.g., a short-range wireless communication network), or may communicate with at least one of an electronic device (104) or a server (108) via a second network (199) (e.g., a long-range wireless communication network). In one embodiment, the electronic device (101) may communicate with the electronic device (104) via the server (108). According to one embodiment, the electronic device (101) may include a processor (120), a memory (130), an input module (150), an audio output module (155), a display module (160), an audio module (170), a sensor module (176), an interface (177), a connection terminal (178), a haptic module (179), a camera module (180), a power management module (188), a battery (189), a communication module (190), a subscriber identification module (196), or an antenna module (197). In some embodiments, the electronic device (101) may omit at least one of these components (e.g., the connection terminal (178)), or may have one or more other components added. In some embodiments, some of these components (e.g., the sensor module (176), the camera module (180), or the antenna module (197)) may be integrated into one component (e.g., the display module (160)).

[0025] The processor (120) may, for example, execute software (e.g., a program (140)) to control at least one other component (e.g., a hardware or software component) of the electronic device (101) connected to the processor (120) and perform various data processing or operations. According to one embodiment, as at least a part of the data processing or operations, the processor (120) may store commands or data received from other components (e.g., a sensor module (176) or a communication module (190)) in a volatile memory (132), process the commands or data stored in the volatile memory (132), and store result data in a non-volatile memory (134). According to one embodiment, the processor (120) may include a main processor (121) (e.g., a central processing unit or an application processor) or an auxiliary processor (123) (e.g., a graphics processing unit, a neural processing unit (NPU), an image signal processor, a sensor hub processor, or a communication processor) that can operate independently or together with the main processor (121). For example, when the electronic device (101) includes the main processor (121) and the auxiliary processor (123), the auxiliary processor (123) may be configured to use less power than the main processor (121) or to be specialized for a given function. The auxiliary processor (123) may be implemented separately from the main processor (121) or as a part thereof.

[0026] The auxiliary processor (123) may control at least a portion of functions or states associated with at least one component (e.g., a display module (160), a sensor module (176), or a communication module (190)) of the electronic device (101), for example, on behalf of the main processor (121) while the main processor (121) is in an inactive (e.g., sleep) state, or together with the main processor (121) while the main processor (121) is in an active (e.g., application execution) state. In one embodiment, the auxiliary processor (123) (e.g., an image signal processor or a communication processor) may be implemented as a part of another functionally related component (e.g., a camera module (180) or a communication module (190)). In one embodiment, the auxiliary processor (123) (e.g., a neural network processing unit) may include a hardware structure specialized for processing artificial intelligence models. The artificial intelligence models may be generated through machine learning. This learning can be performed, for example, on the electronic device (101) itself where the artificial intelligence model is executed, or can be performed through a separate server (e.g., server (108)). The learning algorithm can include, for example, supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning, but is not limited to the examples described above. The artificial intelligence model can include multiple artificial neural network layers.The artificial neural network may be one of a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), a restricted Boltzmann machine (RBM), a deep belief network (DBN), a bidirectional recurrent deep neural network (BRDNN), a deep Q-network, or a combination of two or more of the above, but is not limited to the examples described above. In addition to, or alternatively to, a hardware structure, an artificial intelligence model may include a software structure.

[0027] The memory (130) can store various data used by at least one component (e.g., processor (120) or sensor module (176)) of the electronic device (101). The data can include, for example, software (e.g., program (140)) and input data or output data for commands related thereto. The memory (130) can include volatile memory (132) or non-volatile memory (134).

[0028] The program (140) may be stored as software in the memory (130) and may include, for example, an operating system (142), middleware (144), or an application (146).

[0029] The input module (150) can receive commands or data to be used in a component of the electronic device (101) (e.g., a processor (120)) from an external source (e.g., a user) of the electronic device (101). The input module (150) can include, for example, a microphone, a mouse, a keyboard, a key (e.g., a button), or a digital pen (e.g., a stylus pen).

[0030] The audio output module (155) can output audio signals to the outside of the electronic device (101). The audio output module (155) can include, for example, a speaker or a receiver. The speaker can be used for general purposes, such as multimedia playback or recording playback. The receiver can be used to receive incoming calls. In one embodiment, the receiver can be implemented separately from the speaker or as part of the speaker.

[0031] The display module (160) can visually provide information to an external party (e.g., a user) of the electronic device (101). The display module (160) may include, for example, a display, a holographic device, or a projector, and a control circuit for controlling the device. In one embodiment, the display module (160) may include a touch sensor configured to detect a touch, or a pressure sensor configured to measure the intensity of a force generated by the touch.

[0032] The audio module (170) can convert sound into an electrical signal, or vice versa, convert an electrical signal into sound. According to one embodiment, the audio module (170) can acquire sound through the input module (150), output sound through the sound output module (155), or an external electronic device (e.g., electronic device (102)) (e.g., speaker or headphone) directly or wirelessly connected to the electronic device (101).

[0033] The sensor module (176) can detect the operating status (e.g., power or temperature) of the electronic device (101) or the external environmental status (e.g., user status) and generate an electrical signal or data value corresponding to the detected status. According to one embodiment, the sensor module (176) can include, for example, a gesture sensor, a gyro sensor, a barometric pressure sensor, a magnetic sensor, an acceleration sensor, a grip sensor, a proximity sensor, a color sensor, an IR (infrared) sensor, a biometric sensor, a temperature sensor, a humidity sensor, or an illuminance sensor.

[0034] The interface (177) may support one or more designated protocols that may be used to directly or wirelessly connect the electronic device (101) with an external electronic device (e.g., the electronic device (102)). In one embodiment, the interface (177) may include, for example, a high definition multimedia interface (HDMI), a universal serial bus (USB) interface, an SD card interface, or an audio interface.

[0035] The connection terminal (178) may include a connector through which the electronic device (101) may be physically connected to an external electronic device (e.g., electronic device (102)). According to one embodiment, the connection terminal (178) may include, for example, an HDMI connector, a USB connector, an SD card connector, or an audio connector (e.g., a headphone connector).

[0036] The haptic module (179) can convert electrical signals into mechanical stimuli (e.g., vibration or movement) or electrical stimuli that a user can perceive through tactile or kinesthetic sensations. According to one embodiment, the haptic module (179) can include, for example, a motor, a piezoelectric element, or an electrical stimulation device.

[0037] The camera module (180) can capture still images and videos. According to one embodiment, the camera module (180) may include one or more lenses, image sensors, image signal processors, or flashes.

[0038] The power management module (188) can manage power supplied to the electronic device (101). According to one embodiment, the power management module (188) can be implemented as, for example, at least a part of a power management integrated circuit (PMIC).

[0039] A battery (189) may power at least one component of the electronic device (101). In one embodiment, the battery (189) may include, for example, a non-rechargeable primary battery, a rechargeable secondary battery, or a fuel cell.

[0040] The communication module (190) may support the establishment of a direct (e.g., wired) communication channel or a wireless communication channel between the electronic device (101) and an external electronic device (e.g., electronic device (102), electronic device (104), or server (108)), and the performance of communication through the established communication channel. The communication module (190) may operate independently from the processor (120) (e.g., application processor) and may include one or more communication processors that support direct (e.g., wired) communication or wireless communication. According to one embodiment, the communication module (190) may include a wireless communication module (192) (e.g., a cellular communication module, a short-range wireless communication module, or a global navigation satellite system (GNSS) communication module) or a wired communication module (194) (e.g., a local area network (LAN) communication module, or a power line communication module). Among these communication modules, the corresponding communication module can communicate with an external electronic device (104) via a first network (198) (e.g., a short-range communication network such as Bluetooth, wireless fidelity (WiFi) direct, or infrared data association (IrDA)) or a second network (199) (e.g., a long-range communication network such as a legacy cellular network, a 5G network, a next-generation communication network, the Internet, or a computer network (e.g., a LAN or WAN)). These various types of communication modules can be integrated into a single component (e.g., a single chip) or implemented as multiple separate components (e.g., multiple chips). The wireless communication module (192) can verify or authenticate the electronic device (101) within a communication network such as the first network (198) or the second network (199) by using subscriber information (e.g., an international mobile subscriber identity (IMSI)) stored in the subscriber identification module (196).

[0041] The wireless communication module (192) can support 5G networks and next-generation communication technologies following the 4G network, such as NR access technology (new radio access technology). The NR access technology can support high-speed transmission of high-capacity data (eMBB (enhanced mobile broadband)), minimization of terminal power and connection of multiple terminals (mMTC (massive machine type communications)), or high reliability and low latency (URLLC (ultra-reliable and low-latency communications)). The wireless communication module (192) can support, for example, a high-frequency band (e.g., mmWave band) to achieve a high data transmission rate. The wireless communication module (192) can support various technologies for securing performance in a high-frequency band, such as beamforming, massive multiple-input and multiple-output (MIMO), full dimensional MIMO (FD-MIMO), array antenna, analog beam-forming, or large scale antenna. The wireless communication module (192) can support various requirements specified in the electronic device (101), an external electronic device (e.g., the electronic device (104)), or a network system (e.g., the second network (199)). According to one embodiment, the wireless communication module (192) can support a peak data rate (e.g., 20 Gbps or more) for eMBB realization, a loss coverage (e.g., 164 dB or less) for mMTC realization, or a U-plane latency (e.g., 0.5 ms or less for downlink (DL) and uplink (UL), or 1 ms or less for round trip) for URLLC realization.

[0042] The antenna module (197) can transmit or receive signals or power to or from an external device (e.g., an external electronic device). In one embodiment, the antenna module (197) may include an antenna including a radiator formed of a conductor or a conductive pattern formed on a substrate (e.g., a PCB). In one embodiment, the antenna module (197) may include a plurality of antennas (e.g., an array antenna). In this case, at least one antenna suitable for a communication method used in a communication network, such as the first network (198) or the second network (199), may be selected from the plurality of antennas, for example, by the communication module (190). A signal or power may be transmitted or received between the communication module (190) and an external electronic device via the at least one selected antenna. In some embodiments, in addition to the radiator, another component (e.g., a radio frequency integrated circuit (RFIC)) may be additionally formed as a part of the antenna module (197).

[0043] According to various embodiments, the antenna module (197) may form a mmWave antenna module. In one embodiment, the mmWave antenna module may include a printed circuit board, an RFIC disposed on or adjacent a first side (e.g., a bottom side) of the printed circuit board and capable of supporting a designated high-frequency band (e.g., a mmWave band), and a plurality of antennas (e.g., an array antenna) disposed on or adjacent a second side (e.g., a top side or a side side) of the printed circuit board and capable of transmitting or receiving signals in the designated high-frequency band.

[0044] At least some of the above components can be interconnected and exchange signals (e.g., commands or data) with each other via a communication method between peripheral devices (e.g., a bus, GPIO (general purpose input and output), SPI (serial peripheral interface), or MIPI (mobile industry processor interface)).

[0045] According to one embodiment, commands or data may be transmitted or received between the electronic device (101) and an external electronic device (104) via a server (108) connected to a second network (199). Each of the external electronic devices (102 or 104) may be the same or a different type of device as the electronic device (101). According to one embodiment, all or part of the operations executed in the electronic device (101) may be executed in one or more of the external electronic devices (102, 104, or 108). For example, when the electronic device (101) is to perform a certain function or service automatically or in response to a request from a user or another device, the electronic device (101) may, instead of or in addition to executing the function or service itself, request one or more external electronic devices to perform the function or at least a part of the service. One or more external electronic devices that receive the request may execute at least a portion of the requested function or service, or an additional function or service related to the request, and transmit the result of the execution to the electronic device (101). The electronic device (101) may process the result as is or additionally and provide it as at least a portion of a response to the request. For this purpose, cloud computing, distributed computing, mobile edge computing (MEC), or client-server computing technology may be used, for example. The electronic device (101) may provide an ultra-low latency service by using distributed computing or mobile edge computing, for example. In one embodiment, the external electronic device (104) may include an Internet of Things (IoT) device. The server (108) may be an intelligent server utilizing machine learning and / or a neural network. According to one embodiment, the external electronic device (104) or the server (108) may be included in the second network (199).The electronic device (101) can be applied to intelligent services (e.g., smart home, smart city, smart car, or healthcare) based on 5G communication technology and IoT-related technology.

[0046] FIG. 2 is a block diagram illustrating an integrated intelligence system according to various embodiments.

[0047] Referring to FIG. 2, according to one embodiment, the integrated intelligence system may include an electronic device (210) (e.g., electronic device (101) of FIG. 1), an intelligent server (230) (e.g., server (108) of FIG. 1), and a service server (250) (e.g., server (108) of FIG. 1).

[0048] According to one embodiment, the electronic device (210) may be a terminal device (or electronic device) that can connect to the Internet, for example, a mobile phone, a smart phone, a personal digital assistant (PDA), a laptop computer, a TV, white goods, a wearable device, an HMD, or a smart speaker.

[0049] According to the illustrated embodiment, the electronic device (210) may include a communication interface (213) (e.g., the interface (177) of FIG. 1), a microphone (212) (e.g., the input module (150) of FIG. 1), a speaker (216) (e.g., the audio output module (155) of FIG. 1), a display module (211) (e.g., the display module (160) of FIG. 1), a memory (215) (e.g., the memory (130) of FIG. 1), or a processor (214) (e.g., the processor (120) of FIG. 1). The components listed above may be operatively or electrically connected to each other. The electronic device (210) may include at least some of the configurations and / or functions of the electronic device (101) of FIG. 1.

[0050] In one embodiment, the communication interface (213) may be configured to connect to an external device and transmit and receive data. In one embodiment, the microphone (212) may receive sound (e.g., user speech) and convert it into an electrical signal. In one embodiment, the speaker (216) may output the electrical signal as sound (e.g., voice).

[0051] In one embodiment, the display module (211) may be configured to display an image or video. In one embodiment, the display module (211) may also display a graphical user interface (GUI) of a running app (or application program). In one embodiment, the display module (211) may receive a touch input via a touch sensor. For example, the display module (211) may receive a text input via a touch sensor in an on-screen keyboard area displayed within the display module (211).

[0052] According to one embodiment, the memory (215) may store a client module (218), a software development kit (SDK) (217), and a plurality of apps (219a, 219b). The client module (218) and the SDK (217) may constitute a framework (or solution program) for performing general-purpose functions. In addition, the client module (218) or the SDK (217) may constitute a framework for processing user input (e.g., voice input, text input, touch input).

[0053] According to one embodiment, the plurality of apps (219a, 219b) stored in the memory (215) may be programs for performing a specified function. According to one embodiment, the plurality of apps may include a first app (219a) and a second app (219b). According to one embodiment, each of the plurality of apps (219a, 219b) may include a plurality of operations for performing a specified function. For example, the apps (219a, 219b) may include an alarm app, a message app, and / or a schedule app. According to one embodiment, the plurality of apps (219a, 219b) may be executed by the processor (214) to sequentially execute at least some of the plurality of operations.

[0054] According to one embodiment, the processor (214) can control the overall operation of the electronic device (210). For example, the processor (214) can be electrically connected to a communication interface (213), a microphone (212), a speaker (216), and a display module (211) to perform a designated operation.

[0055] According to one embodiment, the processor (214) may also execute a program stored in the memory (215) to perform a designated function. For example, the processor (214) may execute at least one of the client module (218) or the SDK (217) to perform the following operations for processing user input. The processor (214) may control the operations of a plurality of apps (219a, 219b), for example, through the SDK (217). The following operations described as operations of the client module (218) or the SDK (217) may be operations executed by the processor (214).

[0056] According to one embodiment, the client module (218) can receive user input. For example, the client module (218) can receive a voice signal corresponding to a user utterance detected through the microphone (212). Alternatively, the client module (218) can receive a touch input detected through the display module (211). Alternatively, the client module (218) can receive a text input detected through a keyboard or a visual keyboard. In addition, the client module (218) can receive various forms of user input detected through an input module included in the electronic device (210) or an input module connected to the electronic device (210). The client module (218) can transmit the received user input to the intelligent server (230). The client module (218) can transmit status information of the electronic device (210) together with the received user input to the intelligent server (230). The status information can be, for example, execution status information of an app.

[0057] In one embodiment, the client module (218) may receive a result corresponding to the received user input. For example, the client module (218) may receive a result corresponding to the received user input if the intelligent server (230) can produce a result corresponding to the received user input. The client module (218) may display the received result on the display module (211). Additionally, the client module (218) may output the received result as audio through the speaker (216).

[0058] According to one embodiment, the client module (218) can receive a plan corresponding to the received user input. The client module (218) can display the results of executing multiple operations of the app according to the plan on the display module (211). For example, the client module (218) can sequentially display the results of executing multiple operations on the display module (211) and output audio through the speaker (216). The electronic device (210) can, for another example, display only some results of executing multiple operations (e.g., the result of the last operation) on the display module (211) and output audio through the speaker (216).

[0059] In one embodiment, the client module (218) may receive a request from the intelligent server (230) to obtain information necessary to produce a result corresponding to the voice input. In one embodiment, the client module (218) may transmit the necessary information to the intelligent server (230) in response to the request.

[0060] According to one embodiment, the client module (218) may transmit result information of executing multiple operations according to a plan to the intelligent server (230). The intelligent server (230) may use the result information to confirm that the received user input has been processed correctly.

[0061] In one embodiment, the client module (218) may include a voice recognition module. In one embodiment, the client module (218) may recognize voice inputs that perform limited functions through the voice recognition module. For example, the client module (218) may execute an intelligent app that processes voice inputs to perform organic actions based on a specified input (e.g., "Wake up!").

[0062] According to one embodiment, the intelligent server (230) can receive information related to a user voice input from an electronic device (210) via a communication network. According to one embodiment, the intelligent server (230) can convert data related to the received voice input into text data. According to one embodiment, the intelligent server (230) can generate a plan for performing a task corresponding to the user voice input based on the text data.

[0063] In one embodiment, the plan may be generated by an artificial intelligence (AI) system. The AI ​​system may be a rule-based system, a neural network-based system (e.g., a feedforward neural network (FNN) or a recurrent neural network (RNN)), or a combination of the above or another AI system. In one embodiment, the plan may be selected from a set of predefined plans or may be generated in real time in response to a user request. For example, the AI ​​system may select at least one plan from a plurality of predefined plans.

[0064] According to one embodiment, the intelligent server (230) may transmit the results according to the generated plan to the electronic device (210), or transmit the generated plan to the electronic device (210). According to one embodiment, the electronic device (210) may display the results according to the plan on the display module (211). According to one embodiment, the electronic device (210) may display the results of executing an operation according to the plan on the display module (211).

[0065] According to one embodiment, the intelligent server (230) may include a front end (231), a natural language platform (232), a capsule database (238), an execution engine (233), an end user interface (234), a management platform (235), a big data platform (236), or an analytic platform (237).

[0066] According to one embodiment, the front end (231) can receive user input from the electronic device (210). The front end (231) can transmit a response corresponding to the user input.

[0067] According to one embodiment, the natural language platform (232) may include an automatic speech recognition module (ASR module) (232a), a natural language understanding module (NLU module) (232b), a planner module (232c), a natural language generator module (NLG module) (232d), or a text to speech module (TTS module) (232e).

[0068] According to one embodiment, the automatic speech recognition module (232a) can convert voice input received from the electronic device (210) into text data. According to one embodiment, the natural language understanding module (232b) can use the text data of the voice input to determine the user's intention. For example, the natural language understanding module (232b) can perform syntactic analysis or semantic analysis on user input in the form of text data to determine the user's intention. According to one embodiment, the natural language understanding module (232b) can use linguistic features (e.g., grammatical elements) of morphemes or phrases to determine the meaning of words extracted from the voice input, and can match the meaning of the determined words to the intent to determine the user's intent. The natural language understanding module (223b) can obtain intent information corresponding to the user's utterance. The intent information can be information indicating the user's intent determined by interpreting text data. The intent information can include information indicating an action or function that the user intends to execute using the device.

[0069] According to one embodiment, the planner module (232c) can generate a plan using the intent and parameters determined by the natural language understanding module (232b). According to one embodiment, the planner module (232c) can determine a plurality of domains necessary to perform a task based on the determined intent. The planner module (232c) can determine a plurality of operations included in each of the plurality of domains determined based on the intent. According to one embodiment, the planner module (232c) can determine parameters necessary to execute the determined plurality of operations or result values ​​output by the execution of the plurality of operations. The parameters and the result values ​​can be defined as concepts of a specified format (or class). Accordingly, the plan can include a plurality of operations and a plurality of concepts determined by the user's intent. The planner module (232c) can determine the relationships between the plurality of operations and the plurality of concepts in a stepwise (or hierarchical) manner. For example, the planner module (232c) can determine the execution order of a plurality of actions based on the user's intention based on a plurality of concepts. In other words, the planner module (232c) can determine the execution order of a plurality of actions based on parameters required for the execution of the plurality of actions and results output by the execution of the plurality of actions. Accordingly, the planner module (232c) can generate a plan including association information (e.g., ontology) between the plurality of actions and the plurality of concepts. The planner module (232c) can generate the plan using information stored in a capsule database that stores a set of relationships between concepts and actions.

[0070] According to one embodiment, the natural language generation module (232d) can convert specified information into text format. The information converted into text format may be in the form of natural language speech. According to one embodiment, the text-to-speech module (232e) can convert text-to-speech information into speech information.

[0071] According to one embodiment, some or all of the functions of the natural language platform (232) may also be implemented in the electronic device (210).

[0072] The capsule database may store information about the relationships between multiple concepts and actions corresponding to multiple domains. According to one embodiment, the capsule may include multiple action objects (or action information) and concept objects (or concept information) included in the plan. According to one embodiment, the capsule database may store multiple capsules in the form of a concept action network (CAN). According to one embodiment, the multiple capsules may be stored in a function registry included in the capsule database.

[0073] The capsule database may include a strategy registry that stores strategy information necessary for determining a plan corresponding to a user input. The strategy information may include reference information for determining a single plan when there are multiple plans corresponding to the user input. According to one embodiment, the capsule database may include a follow-up registry that stores information on follow-up actions for suggesting follow-up actions to a user in a given situation. The follow-up actions may include, for example, follow-up utterances. According to one embodiment, the capsule database may include a layout registry that stores layout information of information output through the electronic device (210). According to one embodiment, the capsule database may include a vocabulary registry that stores vocabulary information included in the capsule information. According to one embodiment, the capsule database may include a dialog registry that stores information on dialogue (or interaction) with the user. The capsule database may update stored objects through a developer tool. The developer tool may include, for example, a function editor for updating action objects or concept objects. The developer tool may include a vocabulary editor for updating vocabulary. The developer tool may include a strategy editor for creating and registering strategies that determine plans. The developer tool may include a dialog editor for creating conversations with users.The developer tool may include a follow-up editor that activates follow-up goals and allows editing of follow-up utterances that provide hints. The follow-up goals may be determined based on the currently set goals, user preferences, or environmental conditions. In one embodiment, the capsule database may also be implemented within the electronic device (210).

[0074] In one embodiment, the execution engine (233) can use the generated plan to produce a result. The end user interface (234) can transmit the produced result to the electronic device (210). Accordingly, the electronic device (210) can receive the result and provide the received result to the user. In one embodiment, the management platform (235) can manage information used in the intelligent server (230). In one embodiment, the big data platform (236) can collect user data. In one embodiment, the analysis platform (237) can manage the quality of service (QoS) of the intelligent server (230). For example, the analysis platform (237) can manage the components and processing speed (or efficiency) of the intelligent server (230).

[0075] According to one embodiment, the service server (250) may provide a service (e.g., food ordering or hotel reservation) specified to the electronic device (210). According to one embodiment, the service server (250) may be a server operated by a third party. According to one embodiment, the service server (250) may provide information for generating a plan corresponding to the received voice input to the intelligent server (230). The provided information may be stored in a capsule database. In addition, the service server (250) may provide result information according to the plan to the intelligent server (230). The service server (250) may include a plurality of service providers (e.g., CP Service A (251), CP Service B (252), CP Service C (253)), and each of the service providers (251, 252, 253) may provide a function for a domain associated with each capsule stored in the capsule database (238) of the intelligent server (230).

[0076] In the integrated intelligence system described above, the electronic device (210) can provide various intelligent services to the user in response to user input. The user input may include, for example, input via a physical button, touch input, or voice input.

[0077] According to one embodiment, the electronic device (210) may provide a voice recognition service through an intelligent app (or voice recognition app) stored within the device. In this case, for example, the electronic device (210) may recognize a user utterance or voice input received through the microphone (212) and provide the user with a service corresponding to the recognized voice input.

[0078] According to one embodiment, the electronic device (210) may perform a designated operation based on the received voice input, either alone or together with the intelligent server (230) and / or the service server (250). For example, the electronic device (210) may execute an app corresponding to the received voice input and perform a designated operation through the executed app.

[0079] According to one embodiment, when an electronic device (210) provides a service together with an intelligent server (230) and / or a service server (250), the electronic device (210) may detect a user's speech using the microphone (212) and generate a signal (or voice data) corresponding to the detected user's speech. The electronic device (210) may transmit the voice data to the intelligent server (230) via a network (240) using a communication interface (213).

[0080] In one embodiment, an intelligent server (230) may generate a plan for performing a task corresponding to a voice input received from an electronic device (210), or a result of performing an operation according to the plan, in response to the voice input. The plan may include, for example, a plurality of operations for performing a task corresponding to a user's voice input, and a plurality of concepts related to the plurality of operations. The concept may define parameters input to the execution of the plurality of operations, or result values ​​output by the execution of the plurality of operations. The plan may include association information between the plurality of operations and the plurality of concepts.

[0081] According to one embodiment, the electronic device (210) can receive the response using the communication interface (213). The electronic device (210) can output a voice signal generated within the electronic device (210) to the outside using the speaker (216), or can output an image generated within the electronic device (210) to the outside using the display module (211).

[0082] Although FIG. 2 illustrates an example in which voice recognition, natural language understanding and generation, and result production using a plan of user input received from an electronic device (210) are performed on an intelligent server (230), the various embodiments of the present document are not limited thereto. For example, at least some components of the intelligent server (230) (e.g., natural language platform (232), execution engine (233), capsule database (238)) may be embedded in the electronic device (210) (or the electronic device (101) of FIG. 1), and the operations may be performed by the electronic device (210).

[0083] FIG. 3 is a diagram showing a form in which relationship information between concepts and actions is stored in a database according to various embodiments.

[0084] According to one embodiment, a capsule database (e.g., capsule database (238) of FIG. 2) of an intelligent server (e.g., intelligent server (230) of FIG. 2) may store capsules in the form of a CAN (concept action network) (300). The capsule database may store operations for processing tasks corresponding to a user's voice input and parameters necessary for the operations in the form of a CAN (concept action network).

[0085] According to one embodiment, the capsule database may store a plurality of capsules (capsule (A) (310), capsule (B) (320)) corresponding to each of a plurality of domains (e.g., applications). According to one embodiment, one capsule (e.g., capsule (A) (310)) may correspond to one domain (e.g., location (geo), application). In addition, one capsule may correspond to at least one service provider (e.g., CP 1 (331) or CP 2 (332)) for performing a function for a domain related to the capsule. According to one embodiment, one capsule may include at least one operation (350) and at least one concept (360) for performing a specified function.

[0086] In one embodiment, a natural language platform (e.g., the natural language platform (232) of FIG. 2) can generate a plan for performing a task corresponding to a received speech input using capsules stored in a capsule database. For example, a planner module of the natural language platform (e.g., the planner module (232c) of FIG. 2) can generate a plan using capsules stored in a capsule database. For example, a plan can be generated using actions (311, 313) and concepts (312, 314) of capsule A (310) and actions (321) and concepts (322) of capsule B (320).

[0087] FIG. 4 is a block diagram of a generative artificial intelligence system according to one embodiment.

[0088] Referring to FIG. 4, a generative artificial intelligence system (400) may include a generative AI model (450), an AI framework (440), a user query / response interface (410), an application / service component (430), and a knowledge repository (420). The AI ​​framework (440) may include a prompt design component (442), an API / plugin management component (444), and an output modification component (446).

[0089] According to one embodiment, a user query / response interface (410) may receive a user's input. The user's input may be in the form of natural language, images, and / or videos. In addition, context information may also be transmitted when the user's input is transmitted. The context information may include various additional information at the time of the user's input. For example, the context information may include information related to the user or the electronic device, such as information about the application the user is currently using or information about the user's location. In addition, the user's input may also be in a form that mixes the aforementioned natural language, images, sounds, and context information. In addition, the user's input may also be in a non-natural language form, such as selecting a menu.

[0090] In one embodiment, the user question / response interface (410) may output the results of the generative artificial intelligence system (400) to the user. The output may be in the form of natural language or specific content, and may also be provided in the form of an action requested by the user.

[0091] According to one embodiment, the AI ​​framework (440) can receive user input and coordinate and control each component necessary to perform the user's intention based on the user's query.

[0092] In one embodiment, user input received from the user question / response interface (410) may be transmitted to a prompt design component (442). The prompt design component (442) may be used to generate prompts suitable for inputting the user input into a large language model (LLM) or a large multimodal model (LMM). The prompt design component (442) may be an AI component that uses a machine learning algorithm or a neural network to develop better prompts over time. The prompt design component (442) may access a knowledge repository (420) containing user preference data, a prompt library, and prompt examples based on the user input to obtain and generate prompts, and may transmit the generated prompts to the LLM or LMM.

[0093] In one embodiment, the API / plug-in management component (444) may communicate with external information when there is a request for additional information when passing user input as input to the generative model. The API / plug-in management component (444) may establish a channel for communicating with the outside of the AI ​​interface through the API, and may enable access to various data sources through the established channel. In addition, the API / plug-in management component (444) may request an action through the API that ultimately performs the user input, rather than an intermediate result, when the application or service needs to perform the action. Information obtained from the outside may be used to generate a prompt in the prompt design component (442) together with the user input, or may be passed as input to the generative model.

[0094] In one embodiment, the output modification component (446) (or refiner component) can fine-tune the output from the generative model. For example, the output modification component (446) can verify that the content generated through the LLM and / or LMM is not irrelevant, does not contain biased content, or does not contain harmful content. In addition, the output modification component (446) can determine to what extent the content matches the result desired by the user and, if necessary, can perform additional processing. The output modification component (446) can additionally configure and provide the user with hints to avoid undesired output.

[0095] According to one embodiment, a generative AI model (450) may generally refer to an artificial intelligence neural network that creates new types of data based on user input information. The generative AI model (450) may include a model that generates images and / or a model that generates languages. Representative models for generating images include a generative adversarial network (GAN) and a variational auto encoder (VAE), and examples include a Diffusion-based generative model that uses a VAE and a Transformer structure. A model for generating languages ​​is a model that is trained to statistically output the most appropriate output based on input values, and representative examples include models such as CHAT-GPT 3 and CHAT-GPT 4. In addition, there are also LMMs (large multimodal models) that can recognize various types of data input such as text, images, and voices and generate new data corresponding thereto.

[0096] FIG. 5 illustrates an electronic device and an AI model according to one embodiment.

[0097] According to one embodiment, the AI ​​model (500) (artificial intelligence model) may be configured as a generative AI model, which is an artificial intelligence model that learns various data and generates new information and sentences, and in this document, the generative AI model may be referred to as an AI model (500) or a large language model (LLM). According to one embodiment, the AI ​​model (500) may include an on-device AI model implemented on an electronic device (600) (e.g., the electronic device (101) of FIG. 1, the electronic device (210) of FIG. 2) and a server AI model implemented by an external server (e.g., the intelligent server (230) of FIG. 2), and the server AI model may be implemented and operated on a server of a manufacturer of the electronic device (600), or may be implemented and operated by a third party.

[0098] According to one embodiment, the electronic device (600) may transmit a task request to the AI ​​model (500) based on a user input, which causes the AI ​​model (500) to perform an intended task. For example, when a user of the electronic device (600) inputs a task to be requested from the AI ​​model (500) through a voice input via a microphone or a text input via a keyboard / keypad, the electronic device (600) may generate a prompt including the task request and transmit it to the AI ​​model (500). The prompt is a command for generating a task response in the AI ​​model (500), and may serve to guide the AI ​​model (500) to perform a desired task by the user. The AI ​​model (500) may interpret the prompt received from the electronic device (600), execute the task requested by the user, and generate a result of the executed task as text and / or image information and transmit it to the electronic device (600) as a task response.

[0099] In one embodiment, the AI ​​model (500) may be pre-trained for various types of tasks. There is no specification regarding the training data used to train the AI ​​model (500) and the tasks it can perform.

[0100] According to various embodiments of the present document, the AI ​​model (500) can be trained based on an original document, a result document generated as a result of performing a specific task (e.g., summary, formatting, smart reply, note summary, note reorganization) from the original document, and an evaluation result (or evaluation score) that evaluates the task performance results for the original document and the evaluation document.

[0101] In one embodiment, the electronic device (600) may generate a prompt including an evaluation request for the AI ​​model (500) and transmit it to the AI ​​model (500) to obtain an evaluation result for the resulting document from the AI ​​model (500). The AI ​​model (500) may interpret the prompt received from the electronic device (600), calculate an evaluation result for the resulting document, and transmit it to the electronic device (600) as a response.

[0102] Below, various embodiments that can evaluate a result document using an AI model (500) will be described.

[0103] Figure 6 is a block diagram of an electronic device according to one embodiment.

[0104] Referring to FIG. 6, the electronic device (600) may include a communication module (630), a processor (610), and a memory (620). In various embodiments of the present document, some of the illustrated components may be omitted or replaced with other components. The electronic device (600) may include at least some of the components and / or functions of the electronic device (101) of FIG. 1 and / or the electronic device (210) of FIG. 2. At least some of the components of the illustrated (or not illustrated) electronic device (600) may be operatively, functionally, and / or electrically connected to each other.

[0105] According to one embodiment, the communication module (630) may support wireless communication with an external device using cellular wireless communication (e.g., 4G long term evolution (LTE), 5G new radio (NR)) and / or short-range wireless communication (e.g., Wi-Fi). For example, the electronic device (600) may use the communication module (630) to communicate with an external server (e.g., the intelligent server (230) of FIG. 2) that provides a voice assistant service through a network. The communication module (630) may include at least some of the configurations and / or functions of the communication module (190) of FIG. 1 and / or the communication interface (213) of FIG. 2.

[0106] According to one embodiment, the memory (620) may temporarily or permanently store various data, including volatile memory and non-volatile memory. The memory (620) may include at least a portion of the configuration and / or function of the memory (130) of FIG. 1 and / or the memory (215) of FIG. 2, and may store the program (140) of FIG. 1. The memory (620) may store various applications (e.g., the first app (219a) and the second app (219b) of FIG. 2) and program modules supporting intelligent services (e.g., the client module (218) of FIG. 2).

[0107] According to one embodiment, the memory (620) may store various instructions that may be performed by the processor (610). Such instructions may include control commands such as arithmetic and logical operations, data movement, and / or input / output that may be recognized by the processor (610).

[0108] According to one embodiment, the processor (610) may be configured as one or more processors capable of performing calculations or data processing related to control and / or communication of each component of the electronic device (600). The processor (610) may include at least some of the configurations and / or functions of the processor (120) of FIG. 1 and / or the processor (214) of FIG. 2.

[0109] According to one embodiment, there is no limitation to the computational and data processing functions that the processor (610) can implement on the electronic device (600), but this document will describe various embodiments for evaluating documents using an AI model. The operations of the processor (610), which will be described below, can be performed by loading instructions stored in the memory (620).

[0110] In this document, the description that the processor (610) can perform a certain operation (or function, work, task) may be interpreted to mean substantially the same as that an instruction (or command, computer program) that causes the electronic device (600) (or the processor (610)) to perform the corresponding operation is stored in the memory (620) (e.g., non-volatile memory, storage). In addition, the description that the processor (610) can perform a certain operation may be interpreted to mean substantially the same as that at least one processor (610) can perform the corresponding operation.

[0111] According to one embodiment, the processor (610) may perform a specific task on an original document to generate a result document. Here, the tasks may include, but are not limited to, summarizing, formatting, smart reply, note summarizing, note reorganizing, and automatic formatting. For example, the tasks of this document may include various types of tasks that generate document data in different forms according to predetermined criteria (or algorithms) from document data containing text. Summary is a function that provides a summary of the content contained in the text content, smart reply is a function that recommends a response in response to a received message, note summary is a function that summarizes the content written by a user in a note, note reorganization is a function that rearranges / reorganizes the text of a note written by a user to fit a specified format, and automatic formatting may be a function that sets the format of various contents to an appropriate format. According to one embodiment, the original document and the result document may include text, and may also include other types of data (e.g., images) in addition to text.

[0112] In one embodiment, the resulting document may be generated by a device other than the electronic device (600). For example, the electronic device (600) may obtain a resulting document generated by an AI model, another electronic device, or a server device performing a task on the original document.

[0113] Table 1 below is an example of the original document and the resulting document.

[0114] The most complete village should possess both the villagers as the subjects of transmission and the conditions of a village as a geographical space. In other words, what can be clearly called village folklore is folklore that is transmitted within the village by the villagers as the subjects. Such folklore includes family folklore and household folklore. Since families form homes and live within the village, they possess both the conditions of a villager as a subject and the conditions of a village as a space, and therefore, there is no reason not to call these folklore village folklore. Ultimately, since family folklore or household folklore are also included in the concept of village folklore, the identity of village folklore is once again shaken. Then, not only would the identity of family folklore or household ethnicity not exist, but it would be even more difficult to secure the identity of village folklore and local folklore. {"title": "Identity of Village Folklore""classification": "Lecture note""summary": "The Relationship Between Village Folklore Identity and Family Folklore and Household Folklore"}Generative AI, exemplified by models like Generative Adversarial Networks (GANs), is a subset of artificial intelligence focused on producing new data instances resembling existing datasets. The core concept involves training a generator to create synthetic data and a discriminator to distinguish between real and generated data. This adversarial training process continues iteratively until the generator produces data convincingly similar to real examples.Generative models, such as GANs, have applications in various domains, including image synthesis, style transfer, and text-to-image generation. One of the key achievements of generative AI is its ability to learn patterns and structures from given datasets, enabling the creation of novel content. However, ethical considerations arise due to the potential misuse of generative AI, as seen in the creation of deepfakes. Striking a balance between innovation and responsible use is crucial for the development of generative AI technologies. Continued research in this field holds the promise of creating AI systems that contribute positively to diverse domains while mitigating associated risks.{"title": "Unveiling Generative AI: Applications and Ethical Considerations""Classification": "information""summary": Generative AI, exemplified by models like GANs, creates synthetic data resembling existing datasets, with applications in image synthesis and style transfer, while ethical concerns focus on potential misuse.

[0115] According to one embodiment, the processor (610) may generate the original document and the result document as a CSV (comma-separated values) format file. Here, the CSV format may be a text file in which text fields are separated by commas. The processor (610) may input the original document and the result document by dividing rows and columns into the CSV file. In this way, the generated CSV file may be used to generate a prompt, which will be described later. According to one embodiment, the processor (610) may perform an evaluation operation for a task performed to generate a result document from the original document. According to one embodiment, the processor (610) may determine evaluation items for evaluating the task, request an AI model to evaluate at least some of the evaluation items, and then receive an evaluation result (or a first evaluation result), and for at least some of the other items, the processor (610) may execute a predetermined program code to analyze and directly evaluate the contents of the result document to generate an evaluation result (or a second evaluation result).

[0116] According to one embodiment, the memory (620) may store a database that maps each task type to a plurality of evaluation items. For example, the database stored in the memory (620) may map and store evaluation items for evaluating the result document for each task type, such as summary, format organization, smart reply, note summary, note reorganization, and automatic format setting. For example, the evaluation items for evaluating the result document of note summary, which is an example of a task, may include evaluation items related to the content of the title, the vocabulary of the title, the sentence structure of the title, the content of the summary, the sentence structure of the summary, and the length of the summary.

[0117] According to one embodiment, the database stored in the memory (620) can store at least one of a directive to be included in a prompt, an evaluation method, a type of the evaluation item, and a vector value or weight for each evaluation item by mapping it.

[0118] In one embodiment, the instruction may include textual information that can be entered into a prompt used to request an AI model to provide an evaluation result for a given evaluation item. For example, an instruction may be mapped to each evaluation item corresponding to a specific task, or a single instruction may be mapped to a specific task.

[0119] In one embodiment, the evaluation method (or evaluation subject) may include information regarding whether the evaluation of the corresponding evaluation item will be performed by an AI model or whether the electronic device (600) will directly evaluate the item by executing a predetermined program code. For example, the processor (610) may create a prompt including evaluation items for which the evaluation method is defined as an AI model among the evaluation items and transmit it to the AI ​​model. Evaluation items for which the evaluation method is defined as code may be directly evaluated by executing a predetermined program code.

[0120] In one embodiment, the weight may mean a weight to be multiplied by the evaluation score of the corresponding evaluation item. In one embodiment, the type of the evaluation item defines whether the evaluation item is positive or negative, and the vector value may be assigned a value of 1 if the type of the evaluation item is positive, and a value of -1 if the type of the evaluation item is negative. For example, if the type of a specific evaluation item is positive and the vector value is 1, the evaluation score of the corresponding evaluation item evaluated by the AI ​​model or processor (610) may be multiplied by the weight and 1, and if the type of a specific evaluation item is negative and the vector value is -1, the evaluation score of the corresponding evaluation item evaluated by the AI ​​model or processor (610) may be multiplied by the weight and -1.

[0121] According to one embodiment, the electronic device (600) may store a database that maps at least one of evaluation items, instructions to be included in a prompt, evaluation methods, types of evaluation items, and vector values ​​or weights for each type of task, as shown in Tables 2 to 5 below.

[0122] Table 2 below shows examples of assessment items mapped to note summaries among task types, instructions to include in prompts, assessment methods, types and vector values ​​of assessment items, and weights.

[0123] NoKey Note Summary Qualitative Evaluation Items prompt Method type vector weight 1 Title Content of the title: Does the title include the core keywords of the original text? representation: Is [Title] the title of [Original]? AI positive 192 Title Content of the title: Does the title include content that is not in the original text? inserted: Does [Title] have content that in [Original]? AI positive 103 Title Vocabulary of the title: Does the summary use vocabulary accurately and appropriately? core: Does [Title] and [Orignal] have the same meaning? AI positive 144 Title Sentence structure of the title: Is the summary's sentence structure clear? t_structure: is [Title] structure clear? AI positive 125 Title Grammar of the title: Is the title grammar correct? t_grammar: is the grammar of [Title] correct? AI positive 126 Title Length of the title: The title satisfies the length criteria. Are you satisfied? Code Positive 117 Classification Classification Is the classification appropriate? Category: Is [Classification] properly classified among [Categories]? AI Positive 1108 Summary Summary: Does the summary contain the main content of the original text? labstract: Is (Summary) the summary of [Original]? AI Positive 199 Summary Summary: Does the summary contain content that is not in the original text? contain: Is [Summary] content that in [Original]? AI Positive 1010 Summary Summary: Do the summary and the original text have the same meaning? represent: Does [Summary] and [Original] have the same meaning? Affirmative 1811 Summary Sentence structure: Is the summary's sentence structure clear? s_structure: is[Summary] structure clear? AI Affirmative 1212 Summary Grammar: Is the summary grammatically correct? s_grammar: is the grammar of [Summary] correct? AI Affirmative 1213 Summary Sentence structure: Is the summary easy to understand? readablility: is [Summary] easy-to-understand? AI Affirmative 1214 Summary Sentence structure: Is the summary complete? completed: is [Summary] completed with a complete sentence? AI Affirmative 1415 Summary Vocabulary: Does the summary use vocabulary correctly and appropriately? vocabulary: is [Summary] use vocabulary correctly and appropriately? AI Affirmative 1416 Summary Length: Does the summary satisfy the length requirement? Code Affirmative 11

[0124] Table 3 below shows examples of evaluation items mapped to Smart Reply among task types, instructions to include in prompts, evaluation methods, types and vector values ​​of evaluation items, and weights.

[0125] NoKey Smart Reply Qualitative Evaluation Items promptMethodTypevectorWeight1overall Was the overall conversation considered? Is [Reply] based on the overall context of [Original]? AI positive 11.62 recent Was the last part of the conversation primarily considered? Does [Reply] considered more on the last part of the [Original]? AI positive 11.23 consistency Was the tone of the existing conversation maintained? Does [Reply] maintain the tone of the [Original]? AI positive 11.24 contradiction Was the answer consistent with the facts and numbers mentioned in the conversation? Does [Reply] not contradict the facts and numbers m AI positive 11.75 Is it appropriate as a reply to me's perspective? Does [Reply] replied from "Me" perspective? AI positive 11.76 intene_diversity Was the answer output with diverse intentions? Are 4 [Reply] responses diverse in intent? AI positive 10.37 tone_diversity Are messages output in diverse tones? Are 4 [Reply] responses diverse in tone? AI positive 10.28 count Are 4 messages selected? Code positive 10.1

[0126] Table 4 below shows examples of evaluation items mapped to the automatic format settings for each task type, instructions to include in the prompt, evaluation methods, types and vector values ​​of evaluation items, and weights.

[0127] NoKey Auto Format Setting Qualitative Evaluation Items Prompt Method Type Vector Weight 1 Title Did the title use a long sentence from the original text? Did [Title] use a long sentence from [Orignal]? Code Negative-122 Title Is [Title] one or two words? Code Negative-113 Title Is the title concise? Is [Title] a concise sentence? AI Positive 114 Title Does the title contain the main points of the original text? AI Positive 155 Subheading Is [Subtitle] the same text as [Title]? Code Negative-136 Subheading Did the subtitle use a long sentence from [Detail]? Code Negative-127 Subheading Is the subtitle one or two words? Is [Subtitle] one or two words? Code Negative-118 subheading Is the subheading concise? Is [Subtitle] a concise sentence? AI Positive 119 subheading Does the subheading contain the main points of the details? Does [Subtitle] contain the main points of [Detail]? AI Positive 1510 details Is the details longer than the subheading? Is [Detail] longer than [Subtitle]? Code Positive 1211 details Is the details the same text as the subheading? Is [Detail] the same text as [Subtitle]? Code Negative-1312 details Were the details extracted without summarization in (Original)? AI Positive 1713 details Are the details written in complete sentences? [Detail] areAll the finished sentences? AI Positive 1514 details Does [Detail] contain the content of [Original]? AI Positive 1815 details Is [Detail] almost the same meaning as [Orignal]? AI Positive 1816 Output Is the grammar of [Output] correct? AI Positive 1117 Output Was the vocabulary used correctly and appropriately in [Output]? AI Positive 11

[0128] Table 5 below shows examples of evaluation items mapped to voice recording among task types, instructions to include in prompts, evaluation methods, types and vector values ​​of evaluation items, and weights.

[0129] NoVoice recordingQualitative evaluation itemsPromptMethodTypeVectorWeight1Evaluate whether all items in the summary are composed of important content in the original text. AI Positive122Evaluate whether there are no typos and the grammar is correct. AI Positive10.23Evaluate whether the result has almost the same meanings as the original content. AI Positive11.64Evaluate whether the title of the result is concise and main point of the original text. AI Positive115Evaluate whether keywords represent title and summary well. AI Positive116Evaluate whether the summary of the result and the original are almost same in principle. AI Positive127Evaluate whether the summary summarizes the original text in a simple sentence. Assess whether the result summarizes the original into a simple sentence. AI Positive 118 Assess whether there are two keywords. Code Positive 10.49 Assess whether the keywords are included in the summary and title. Code Positive 10.410 Assess whether the summary has three items. Code Positive 10.4

[0130] According to one embodiment, when an evaluation is triggered for an original document and a result document, the processor (610) may check the type of task performed to generate the corresponding result document. For example, the processor (610) may check whether the result document was generated by performing any one of the following tasks on the original document: summary, formatting, smart reply, note summary, note reorganization, and automatic formatting. According to one embodiment, the result document may be generated as a JSON (Java Script Object Notation) file, and when performing the task to create the result document as a JSON file, the type of the performed task may be defined by a tag value. The processor (610) may check the tag value of the result document file to check the type of the task. According to another embodiment, the processor (610) may analyze the contents of the original document and the result document to determine whether any task was performed on the result document. Alternatively, the processor (610) may generate a prompt including instructions requesting the AI ​​model to confirm the original document, the resulting document, and what task was performed, and transmit the prompt to the AI ​​model, and may confirm what task was performed on the resulting document based on the response of the AI ​​model.

[0131] According to one embodiment, the processor (610) can check the evaluation items mapped to the type of the performed task in the database stored in the memory (620). For example, if the tag value of the result document or the analysis result or the performed task is confirmed as a note summary, at least one of the evaluation items stored in the database as shown in Table 2 and the instructions to be included in the prompts mapped and stored for each evaluation item, the evaluation method, the type and vector value of the evaluation item, or the weight can be checked.

[0132] In one embodiment, the processor (610) may generate a prompt to request an evaluation of the task performed on the resulting document by the AI ​​model. The prompt may include at least some of the original document, the resulting document, and the determined evaluation items.

[0133] According to one embodiment, the processor (610) may check the evaluation items mapped to the task performed on the result document in the database of the memory (620), and may generate a prompt using at least some of the evaluation items, instructions, types and vectors of the evaluation items, or weights of the first set of evaluation items for which the evaluation method is designated as an AI model among the identified evaluation items. For example, if the text performed on the result document is a note summary, the evaluation items, instructions to be included in the prompt, the evaluation method, the types and vector values ​​of the evaluation items, and the weights may be mapped and stored as shown in Table 2, and the processor (610) may check the evaluation items for which the evaluation method is designated as an AI model among these as the first set of evaluation items.

[0134] Examples of prompts generated by the processor (610) and responses generated from the AI ​​model will be described in more detail with reference to FIGS. 11 and 12.

[0135] In one embodiment, the processor (610) may transmit the generated prompt to the AI ​​model. For example, if the evaluation is performed by a server AI model, the processor (610) may transmit the generated prompt to the server AI model via a network using the communication module (630). Alternatively, if the evaluation is performed by an on-device AI model, the processor (610) may transmit the generated prompt to an AI processor configured independently of the processor (610) or an AI module including a program executable by the processor (610).

[0136] According to one embodiment, an AI model (e.g., an on-device AI model, a server AI model) may evaluate a task performed on a result document based on a prompt transmitted from an electronic device (600) and generate an evaluation result. According to one embodiment, the AI ​​model may be trained based on an original document, the original document, the result document, and the evaluation result, and accordingly, the AI ​​model may calculate an evaluation score for each evaluation item when the original document, the result document, and the evaluation items are transmitted through a prompt. The AI ​​model may check each evaluation item in the prompt, evaluate the result document for each evaluation item, and calculate the evaluation score as a value within a predetermined range (e.g., 0 to 1). The AI ​​model may transmit the determined evaluation score for each evaluation item to the electronic device (600) as a response to the prompt. In this document, the evaluation result generated by the AI ​​model may also be referred to as a first evaluation result.

[0137] According to one embodiment, the processor (610) may receive a response including a first evaluation result from the AI ​​model.

[0138] According to one embodiment, the processor (610) may analyze the contents of the result document to obtain a second evaluation result for the task performed to generate the result document. For example, the electronic device (600) may store a program code capable of analyzing and evaluating the contents of the result document, and the processor (610) may execute the program code to generate an evaluation result. The processor (610) may evaluate the result document for each evaluation item and calculate an evaluation score as a value within a predetermined range (e.g., 0 to 1). In this document, the evaluation result generated by the processor (610) using the predetermined program code may also be referred to as a second evaluation result.

[0139] According to one embodiment, the processor (610) may generate a second evaluation result by using at least some of the evaluation items, types and vectors of the evaluation items, or weights of the second evaluation item set for which the evaluation method is designated as a code among the evaluation items mapped to the task performed on the corresponding result document. For example, among the evaluation items mapped to the note summary described through Table 2, evaluation items such as the length of the title and the length of the summary may be evaluated by a program code implementing a predetermined algorithm in terms of accuracy and / or resources rather than by an AI model. Accordingly, the database stored in the memory (620) may designate at least some of the evaluation items as an AI model for an evaluation method, and at least some of the evaluation items as a code, in consideration of accuracy and / or resources among the evaluation items mapped to each task.

[0140] In one embodiment, the processor (610) may calculate a final evaluation result based on the first evaluation result received from the AI ​​model and the second evaluation result generated by the program code. For example, the processor (610) may multiply the evaluation score, vector value, and weight for each evaluation item, and add the values ​​obtained for all evaluation items to calculate the final evaluation result for the task performed on the corresponding result document.

[0141] The following mathematical expression 1 shows an example of a formula for calculating the final evaluation result.

[0142]

[0143] In the above mathematical expression 1, S i is the evaluation score evaluated by the AI ​​model or processor (610) for each evaluation item, and W i is the weight of each evaluation item, and n can be the number of evaluation items.

[0144] Table 6 below shows examples of evaluation scores and final evaluation results for each evaluation item.

[0145] Classification Evaluation Items GoodGood 70%Bad 30%Good 50%Bad 50%BadTitleContents of the title: Does the title include the core keywords of the original text? 91.893.1050Contents of the title: Does the title include content that is not in the original text? 0000Vocabulary of the title: Does the vocabulary in the summary use accurate and appropriate vocabulary? 3.20.620.420Sentence structure of the title: Is the sentence structure of the summary clear? 221.960.5Grammar of the title: Is the grammar of the title correct? 2220.3Length of the title: Does the title satisfy the length condition? 10.850.90.8ClassificationIs the classification appropriate? 9.658.757.63.5SummaryContents of the summary: Does the summary include the core content of the original text? 6.481.2152.070Contents of the summary: Does the summary include content that is not in the original text? Summary Content: Does the summary have the same meaning as the original text? 6.08 1.16 1.88 Sentence Structure of the Summary: Does the summary have a clear sentence structure? 221.47 0.25 Grammar of the Summary: Is the summary grammatically correct? 221.48 0.25 Sentence Structure of the Summary: Is the summary easy to understand? 1.99 1.98 1.47 0.25 Sentence Structure of the Summary: Is the summary complete with complete sentences? 44 30.5 Vocabulary of the Summary: Does the summary use vocabulary accurately and appropriately? 44 30.5 Length of the Summary: Does the summary meet the length requirements? 10.05 0.10.35 Final Evaluation Results 4.6 33.17 3.03 1.48

[0146] The electronic device (600) can improve many limitations that occurred when a person manually evaluated the result document by using the AI ​​model as described above. Instructions for performing the operations of the electronic device (600) (or processor (610)) described above can be stored in a computer-readable recording medium. The recording medium can be tangible and non-transitory. The recording medium can store one or more computer programs including the instructions.

[0147] Figure 7 is a block diagram of a document evaluation system according to one embodiment.

[0148] In one embodiment, the document evaluation system (700) can evaluate a result document (714) generated by performing a specific task on an original document (712). Here, the task may include, but is not limited to, summarizing, formatting, smart replying, note summarizing, note reorganizing, and automatic formatting. In one embodiment, the result document (714) may be generated by an electronic device (e.g., the electronic device (600) of FIG. 6), or an AI model, another electronic device, or a server device may perform a task on the original document (712) to generate the result document (714).

[0149] According to one embodiment, the document feature recognition module (720) can check the type of task performed to generate the result document (714) when an evaluation is triggered for the original document (712) and the result document (714). According to one embodiment, the result document (714) can be generated as a json (java script object notation) file, and the document feature recognition module (720) can check the type of task defined in the tag value of the json file. Alternatively, the document feature recognition module (720) can analyze the contents of the original document (712) and the result document (714) to determine whether any task was performed on the result document (714).

[0150] According to one embodiment, the document feature recognition module (720) can obtain at least some of the evaluation items corresponding to the type of the identified task and the instructions to be included in the mapped prompt for each evaluation item, the evaluation method, the type and vector value of the evaluation item, and the weight. For example, the electronic device can map and store each type of task and a plurality of evaluation items, and at least some of the instructions, the evaluation method, the type and vector value of the evaluation item, and the weight for each evaluation item, and the document feature recognition module can obtain values ​​corresponding to the type of the identified task from the database.

[0151] According to one embodiment, the prompt generation module (740) can generate a prompt to request the AI ​​model (500) to evaluate a task performed on the result document (714). The prompt generation module (740) can obtain, as a base prompt (730), evaluation items (732) and instructions (734) corresponding to the identified task. The prompt generation module (740) can generate a prompt that includes at least some of the original document (712), the result document (714), the instructions (734), the evaluation items (732), the type and vector of each evaluation item, or the weight.

[0152] According to one embodiment, the prompt generation module (740) may generate a prompt using at least some of the evaluation items, instructions, types and vectors of the evaluation items, or weights of the first evaluation item set, for which the evaluation method is designated as an AI model (500), among the evaluation items corresponding to the task performed on the result document (714) identified in the database.

[0153] In one embodiment, the prompt generation module (740) may transmit the generated prompt to the AI ​​model (500). In one embodiment, the AI ​​model (500) may check each evaluation item in the prompt, evaluate the result document (714) for each evaluation item, and produce the first evaluation result.

[0154] According to one embodiment, the internal evaluation module (750) may analyze the contents of the result document (714) to generate a second evaluation result for the task performed to generate the result document (714). For example, the internal evaluation module (750) may generate the second evaluation result using a program code that is executed by a processor of an electronic device and that can analyze and evaluate the contents of the result document (714). The internal evaluation module (750) may generate the second evaluation result using at least some of the evaluation items, instructions, types and vectors of the evaluation items, or weights of the second evaluation item set, in which an evaluation method is designated by code, among the evaluation items corresponding to the task performed on the result document (714) identified in the database.

[0155] According to one embodiment, the final evaluation module (760) can obtain the first evaluation result produced by the AI ​​model and the second evaluation result produced by the internal evaluation module. Furthermore, the final evaluation module (760) can obtain weights (735) for each evaluation item.

[0156] In one embodiment, the final evaluation module (760) may calculate a final evaluation result based on the first and second evaluation results. For example, the final evaluation module (760) may multiply the evaluation score, vector value, and weight for each evaluation item, and add the values ​​obtained for all evaluation items to calculate the final evaluation result for the task performed on the corresponding result document.

[0157] FIG. 8 illustrates a method for evaluating the performance results of a task from a source document and a result document according to one embodiment.

[0158] According to one embodiment, the original document (812) may be a text document that is the target of a task, and the result document (814) may be a text document generated as a result of performing the task on the original document (812).

[0159] According to one embodiment, when an evaluation is triggered for an original document and a result document, the electronic device can recognize the type of task performed to generate the result document (820). According to one embodiment, the result document can be generated as a JSON (Java Script Object Notation) file, and the electronic device can identify the type of task defined in the tag value of the JSON file. Alternatively, the electronic device can analyze the contents of the original document and the result document to determine whether a task has been performed on the result document.

[0160] In one embodiment, the electronic device can identify evaluation items mapped to the type of task performed in a database stored in memory. Furthermore, the electronic device can further identify at least one of the mapped instructions, evaluation methods, evaluation item types, vector values, or weights for each identified evaluation item.

[0161] In one embodiment, a prompt may be generated to request an evaluation of a task performed on a result document by an AI model (830). In one embodiment, the electronic device may generate the prompt using at least some of the evaluation items, instructions, types and vectors of evaluation items, or weights of a first set of evaluation items, among the identified evaluation items, for which an evaluation method is designated by the AI ​​model, and transmit the prompt to the AI ​​model.

[0162] In one embodiment, based on the prompts received by the AI ​​model, an evaluation score can be calculated by comparing the original document with the resulting document for each evaluation item, and a first evaluation result can be generated. The AI ​​model can transmit the generated first evaluation result to an electronic device (840).

[0163] According to one embodiment, the electronic device may execute program code capable of analyzing and evaluating the contents of a result document, and by analyzing the contents of the result document, obtain a second evaluation result for the task performed to generate the result document. For example, the electronic device may generate the second evaluation result using at least some of the evaluation items, types and vectors of the evaluation items, or weights of the second evaluation item set, for which the evaluation method is designated by code, among the evaluation items mapped to the task performed on the corresponding result document.

[0164] According to one embodiment, the electronic device may calculate a final evaluation result based on the first evaluation result received from the AI ​​model and the second evaluation result generated by the program code (860). For example, the electronic device may multiply the evaluation score, vector value, and weight for each evaluation item, and add the values ​​obtained for all evaluation items to calculate the final evaluation result for the task performed on the corresponding result document. The electronic device may then apply a weight to the evaluation score for each evaluation item and convert it into a 4-point scale score to generate the final evaluation result.

[0165] Figure 9 illustrates an evaluation item database according to one embodiment.

[0166] According to one embodiment, an electronic device (e.g., the electronic device (600) of FIG. 6) may store a database (910) that maps each task type and a plurality of evaluation items in a memory (e.g., the memory (620) of FIG. 6). For example, the database (910) stored in the memory may map and store evaluation items for evaluating a result document for each task type, such as a summary, format organization, smart reply, note summary, note reorganization, and automatic format setting. According to one embodiment, the database (910) may map and store at least one of a directive to be included in a prompt corresponding to the task (or each evaluation item), an evaluation method, a type and vector value of the evaluation item, or a weight.

[0167] According to one embodiment, a result document generated by executing a specific task on an original document may be generated as a json (java script object notation) file, and when performing a task to create a result document as a json file, the type of the performed task may be defined as a tag value.

[0168] According to one embodiment, the electronic device can check the tag value of the result document to determine the type of task performed on the result document. Referring to FIG. 9, the electronic device can determine that the type of task performed is note summary through the tag value in the first result document (922), that the type of task performed is smart reply through the tag value in the second result document (924), and that the type of task performed is note reorganization through the tag value in the third result document (926).

[0169] According to one embodiment, the electronic device can check the evaluation items mapped to the type of task performed in the database (910) and the instructions mapped to the corresponding task (or each evaluation item).

[0170] Referring to FIG. 9, the electronic device can check the evaluation items and instructions (932) corresponding to the task type of note summary of the first result document (922), check the evaluation items and instructions (934) corresponding to the task type of smart reply of the second result document (924), and check the evaluation items and instructions (936) corresponding to the task type of note reconstruction of the third result document (926).

[0171] In one embodiment, the electronic device may generate prompts to be input into the AI ​​model using the evaluation items and instructions (932, 934, 936) identified in the database (910) to evaluate each result document.

[0172] Figure 10 illustrates information entered into a prompt for document evaluation according to one embodiment.

[0173] In one embodiment, an electronic device (e.g., electronic device (600) of FIG. 6) may generate a prompt (1050) to request an evaluation of a task performed on a result document (1040).

[0174] According to one embodiment, the electronic device may identify the type of task executed to generate the result document (1040) to generate the prompt (1050), and identify evaluation items (1020) corresponding to the type of task identified in a database (e.g., database (910) of FIG. 9) and instructions (1020) mapped and stored to each of the evaluation items (1020), the type and vector of the evaluation items, or at least some of the weights.

[0175] According to one embodiment, the electronic device may generate a prompt (1050) using at least some of the evaluation items, instructions (1020), types and vectors of the evaluation items, or weights of the first set of evaluation items for which an evaluation method is designated as an AI model among the identified evaluation items (1020).

[0176] Referring to FIG. 10, an electronic device may generate a prompt (1050) including an original document (1030), a result document (1040), evaluation items (1020) corresponding to a task identified in a database, and instructions (1020) corresponding to the task. The electronic device may transmit the generated prompt (1050) to an AI model (e.g., an on-device AI model, a server AI model) and receive a response including an evaluation result (e.g., a first evaluation result) from the AI ​​model.

[0177] Figure 11 illustrates a prompt generated for document evaluation according to one embodiment.

[0178] In one embodiment, an electronic device (e.g., electronic device (600) of FIG. 6 ) may generate a prompt to request an evaluation of a task performed on a result document. Referring to FIG. 11 , the prompt may include instructions (1130, 1140), evaluation items (1150), a source document (1110), and a result document (1120).

[0179] According to one embodiment, the original document (1110) may be a text document that is the target of a task, and the result document (1120) may be a text document generated as a result of performing the task on the original document (1110). In the example of FIG. 11, when a note summary for the original document (1110) is generated as the result document (1120), a prompt is illustrated for requesting an AI model to evaluate the task performed on the result document (1120).

[0180] According to one embodiment, the electronic device can determine that the type of the task is a note summary by analyzing the tag value included in the result document (1120) or the original document and the result document, and can check the evaluation items (1150) mapped to the note summary in a database (e.g., the database (910) of FIG. 9). For example, the evaluation items mapped to the note summary can include various evaluation items related to the content of the title, the vocabulary of the title, the sentence structure of the title, the appropriateness of the classification, the content of the summary, the sentence structure of the summary, and the grammar of the summary.

[0181] In one embodiment, the electronic device can check the database for mapped instructions (1130, 1140) for the task and include them in the prompt. Referring to FIG. 11, the categories of the instructions may include categories that can be classified based on the content of the task note summary. For example, the AI ​​model can evaluate whether a category defined in the resulting document is appropriately selected as one of the categories of the instruction.

[0182] In one embodiment, the instructions (1130, 1140) may include one-shot examples of evaluations of the result document. Referring to FIG. 11, the electronic device may specify examples of evaluation scores for each evaluation item in the prompt, with an arbitrary value between 0 and 1. The AI ​​model may evaluate the result document (1120) by referring to the contents of the instructions (1130, 1140) of the prompt and output an evaluation score with a value between 0 and 1 for each evaluation item, as in the examples of evaluation scores included in the instructions (1130, 1140). In addition, the electronic device may include a description of each evaluation item in the prompt.

[0183] Figure 12 illustrates a response of an AI model according to one embodiment.

[0184] In one embodiment, an AI model (e.g., an on-device AI model, a server AI model) can evaluate a task performed on a result document based on a prompt received from an electronic device and generate an evaluation result. In one embodiment, the AI ​​model can be trained based on the original document, the original document, the result document, and the evaluation result, and thus, when the original document, the result document, and the evaluation items are transmitted through the prompt, the AI ​​model can calculate an evaluation score for each evaluation item.

[0185] In one embodiment, the AI ​​model may transmit evaluation results, including evaluation scores calculated for each evaluation item, to the electronic device as a response to the prompt.

[0186] Referring to FIG. 12, the response transmitted from the AI ​​model to the electronic device may include an evaluation score (e.g., result: 0.9. 0.0. 0.75. 0.3. 1.0. 0.2. 0.5. 0.44. 0.95. 0.17. 0.7. 0.33. 1.0. 0.82. 0.72. 0.31) for each evaluation item (e.g., representation, inserted, core, t_structure, t_grammar, t_langauge, category, abstract, contain, represent).

[0187] In one embodiment, the electronic device can determine the final evaluation result based on the first evaluation result received from the AI ​​model and the second evaluation result generated by the program code on the electronic device. For example, the processor can multiply the evaluation score, vector value, and weight for each evaluation item and add the values ​​obtained for all evaluation items to calculate the final evaluation result for the task performed on the corresponding result document.

[0188] An electronic device according to various embodiments of the present document may include a memory and at least one processor operatively connected to the memory.

[0189] According to one embodiment, the memory may store instructions that are executable by at least one processor and, when executed, cause the electronic device to obtain an original document and a result document generated by performing a task on the original document, determine evaluation items for evaluating the task performed to generate the result document, generate a prompt for requesting an AI model to evaluate the task, including at least some of the original document, the result document, and the determined evaluation items, transmit the generated prompt to the AI ​​model, and obtain a first evaluation result for the performed task from the AI ​​model.

[0190] According to one embodiment, the memory may store a database comprising a plurality of instructions including text information inputtable to the prompt and a request to the AI ​​model.

[0191] According to one embodiment, the memory may include instructions causing the electronic device to select a directive corresponding to a task performed to generate the result document from among the directives stored in the database, and to generate the prompt including the directive.

[0192] According to one embodiment, the memory may store instructions that cause the electronic device to analyze the contents of the result document and obtain a second evaluation result for a task performed to generate the result document.

[0193] According to one embodiment, the memory may store instructions that cause the electronic device, in the case of a first set of evaluation items among the evaluation items, to obtain the first evaluation result from the AI ​​model by including the first set of evaluation items in the prompt, and in the case of a second set of evaluation items, to analyze the contents of the result document to obtain the second evaluation result.

[0194] According to one embodiment, the memory may store instructions that cause the electronic device to determine a final evaluation result based on the first evaluation result and the second evaluation result.

[0195] According to one embodiment, the memory may store instructions that cause the electronic device to multiply an evaluation score for each evaluation item in the first evaluation result and the second evaluation result by a weight corresponding to each evaluation item, and determine the final evaluation result based on a sum of the product of the evaluation scores and the weights.

[0196] According to one embodiment, the database stored in the memory stores a mapping between the type of each task and a plurality of evaluation items, and the memory can store instructions that cause the electronic device to check, from the database, evaluation items corresponding to the type of task performed to generate the result document.

[0197] According to one embodiment, the memory may store instructions that cause the electronic device to further verify at least one of a directive, a type of an evaluation item, or a weight mapped to each of the identified evaluation items stored in the database.

[0198] According to one embodiment, the memory may store instructions that cause the electronic device to determine the type of task performed to generate the result document through a tag value included in the result document.

[0199] In one embodiment, the task may include the ability to generate a summary for the original document.

[0200] In one embodiment, the task may include a formatting function for the original document.

[0201] A method performed by an electronic device according to various embodiments of the present document may include an operation of obtaining an original document and a result document generated by performing a task on the original document, an operation of determining evaluation items for evaluating the task performed to generate the result document, an operation of generating a prompt for requesting an AI model to evaluate the task, the prompt including at least some of the original document, the result document, and the determined evaluation items, an operation of transmitting the generated prompt to the AI ​​model, and an operation of obtaining a first evaluation result for the performed task from the AI ​​model.

[0202] According to one embodiment, the electronic device may store a database comprising a plurality of instructions including inputtable text information for a prompt and a request for the AI ​​model.

[0203] According to one embodiment, the operation of generating the prompt may include the operation of selecting a directive corresponding to a task performed to generate the result document from among the directives stored in the database, and the operation of generating the prompt including the directive.

[0204] According to one embodiment, the method may further include an operation of analyzing the contents of the result document to obtain a second evaluation result for a task performed to generate the result document.

[0205] According to one embodiment, the method may include, for a first set of evaluation items among the evaluation items, an operation of obtaining the first evaluation result from the AI ​​model by including the first set of evaluation items in the prompt, and, for a second set of evaluation items, an operation of obtaining the second evaluation result by analyzing the content of the result document.

[0206] According to one embodiment, the method may further include an operation of determining a final evaluation result based on the first evaluation result and the second evaluation result.

[0207] According to one embodiment, the database stores a mapping between a type of each task and a plurality of evaluation items, and the method may include an operation of checking, from the database, evaluation items corresponding to a type of task performed to generate the result document.

[0208] According to one embodiment, the method may further include an operation of confirming the type of task performed to generate the result document through a tag value included in the result document.

[0209] A computer-readable non-transitory recording medium according to various embodiments of the present document may store instructions for performing an operation of obtaining an original document and a result document generated by performing a task on the original document, an operation of determining evaluation items for evaluating the task performed to generate the result document, an operation of generating a prompt for requesting an AI model to evaluate the task, the prompt including at least some of the original document, the result document, and the determined evaluation items, an operation of transmitting the generated prompt to the AI ​​model, and an operation of obtaining a first evaluation result for the performed task from the AI ​​model.

[0210] Electronic devices according to the various embodiments disclosed in this document may take various forms. Electronic devices may include, for example, portable communication devices (e.g., smartphones), computer devices, portable multimedia devices, portable medical devices, cameras, wearable devices, or home appliances. Electronic devices according to the embodiments of this document are not limited to the aforementioned devices.

[0211] The various embodiments of this document and the terminology used therein are not intended to limit the technical features described in this document to specific embodiments, but should be understood to include various modifications, equivalents, or substitutes of the embodiments. In connection with the description of the drawings, similar reference numerals may be used for similar or related components. The singular form of a noun corresponding to an item may include one or more of the items, unless the context clearly indicates otherwise. In this document, each of the phrases "A or B", "at least one of A and B", "at least one of A or B", "A, B, or C", "at least one of A, B, and C", and "at least one of A, B, or C" can include any one of the items listed together in the corresponding phrase among those phrases, or all possible combinations thereof. Terms such as "first," "second," or "first" or "second" may be used merely to distinguish one component from another, and do not limit the components in any other respect (e.g., importance or order). When a component (e.g., a first component) is referred to as "coupled" or "connected" to another (e.g., a second component), with or without the terms "functionally" or "communicatively," it means that the component can be connected to the other component directly (e.g., wired), wirelessly, or through a third component.

[0212] The term "module" used in various embodiments of this document may include a unit implemented in hardware, software, or firmware, and may be used interchangeably with terms such as logic, logic block, component, or circuit. A module may be an integral component, or a minimum unit or part of such a component that performs one or more functions. For example, according to one embodiment, a module may be implemented in the form of an application-specific integrated circuit (ASIC).

[0213] Various embodiments of the present document may be implemented as software (e.g., a program (140)) including one or more instructions stored in a storage medium (e.g., an internal memory (136) or an external memory (138)) readable by a machine (e.g., an electronic device (101)). For example, a processor (e.g., a processor (120)) of the machine (e.g., an electronic device (101)) may call at least one instruction among the one or more instructions stored from the storage medium and execute it. This enables the machine to operate to perform at least one function according to the at least one called instruction. The one or more instructions may include code generated by a compiler or code executable by an interpreter. The machine-readable storage medium may be provided in the form of a non-transitory storage medium. Here, 'non-transitory' simply means that the storage medium is a tangible device and does not contain signals (e.g., electromagnetic waves), and the term does not distinguish between cases where data is stored semi-permanently or temporarily on the storage medium.

[0214] According to one embodiment, the method according to various embodiments disclosed in this document may be provided as a computer program product. The computer program product may be traded between sellers and buyers as a product. The computer program product may be distributed in the form of a device-readable storage medium (e.g., compact disc read-only memory (CD-ROM)) or may be provided through an application store (e.g., Play Store). TM ) or directly between two user devices (e.g., smart phones), online distribution (e.g., downloading or uploading). In the case of online distribution, at least a portion of the computer program product may be at least temporarily stored or temporarily created in a machine-readable storage medium, such as the memory of a manufacturer's server, an application store's server, or an intermediary server.

[0215] According to various embodiments, each component (e.g., a module or a program) of the above-described components may include one or more entities, and some of the entities may be separately arranged in other components. According to various embodiments, one or more components or operations of the aforementioned components may be omitted, or one or more other components or operations may be added. Alternatively or additionally, a plurality of components (e.g., a module or a program) may be integrated into a single component. In such a case, the integrated component may perform one or more functions of each of the plurality of components identically or similarly to those performed by the corresponding component among the plurality of components prior to the integration. According to various embodiments, the operations performed by a module, program, or other component may be executed sequentially, in parallel, iteratively, or heuristically, or one or more of the operations may be executed in a different order, omitted, or one or more other operations may be added.

Claims

1. In electronic devices, memory; and comprising at least one processor operatively connected to said memory; The above memory is executable by at least one processor, and when executed, the electronic device: Obtain the original document and the resulting document generated by performing a task on the original document, Determine evaluation items to evaluate the tasks performed to generate the above result document, Generate a prompt for requesting an AI model to evaluate the task, including at least some of the original document, the result document, and the determined evaluation items, Passing the generated prompt to the AI model, and An electronic device storing instructions for obtaining a first evaluation result for the performed task from the AI model.

2. In paragraph 1, The above memory is, An electronic device storing a database comprising a plurality of instructions including inputtable text information for a prompt and a request for said AI model.

3. In paragraph 2, The above memory, the electronic device, Selecting a directive corresponding to a task performed to generate the result document from among the directives stored in the above database, and An electronic device comprising instructions for generating a prompt including the above directive.

4. In any one of paragraphs 1 to 3, The above memory, the electronic device, An electronic device storing instructions for analyzing the contents of the above result document to obtain a second evaluation result for the task performed to generate the above result document.

5. In paragraph 4, The above memory, the electronic device, For the first set of evaluation items among the above evaluation items, the first evaluation result is obtained from the AI model by including the first set of evaluation items in the above prompt, and for the second set of evaluation items, the second evaluation result is obtained by analyzing the content of the result document. An electronic device storing instructions for determining a final evaluation result based on the first evaluation result and the second evaluation result.

6. In paragraph 5, The above memory, the electronic device, An electronic device storing instructions for multiplying the evaluation score for each evaluation item by a weight corresponding to each evaluation item in the first evaluation result and the second evaluation result, and determining the final evaluation result based on the sum of the product of the evaluation scores and the weights.

7. In any one of paragraphs 1 to 6, The database stored in the above memory stores the type of each task and multiple evaluation items by mapping them, The above memory, the electronic device, An electronic device storing instructions for checking evaluation items corresponding to the type of task performed to generate the above result document from the database.

8. In paragraph 7, The above memory, the electronic device, An electronic device storing instructions for further checking at least one of a mapped instruction, a type of evaluation item, or a weighting of each of the confirmed evaluation items stored in the database.

9. In any one of paragraphs 1 to 8, The above memory, the electronic device, An electronic device storing instructions for identifying the type of task performed to generate the result document through a tag value included in the result document.

10. In any one of paragraphs 1 to 9, An electronic device wherein the task comprises at least one of a function of generating a summary for the original document or a function of formatting the original document.

11. In a method performed by an electronic device, An action to obtain an original document and a result document generated by performing a task on said original document; An action for determining evaluation items for evaluating the tasks performed to generate the above result document; An action of generating a prompt for requesting an AI model to evaluate the task, the action including at least some of the original document, the result document, and the determined evaluation items; An action of transmitting the generated prompt to the AI model; and A method comprising an action of obtaining a first evaluation result for the performed task from the AI model.

12. In paragraph 11, The above electronic device, Store a database containing a plurality of directives that include inputtable text information for the prompt and that include requests to the AI model; The action that generates the above prompt is: An operation of selecting a directive corresponding to a task performed to generate the result document from among the directives stored in the database; and A method comprising generating a prompt including the above directive.

13. In paragraph 12, For the first set of evaluation items among the above evaluation items, an operation of obtaining the first evaluation result from the AI model by including the first set of evaluation items in the prompt, and for the second set of evaluation items, an operation of obtaining the second evaluation result by analyzing the content of the result document; and A method including an operation of determining a final evaluation result based on the first evaluation result and the second evaluation result.

14. In any one of paragraphs 11 to 13, The above database stores the mapping of each task type and multiple evaluation items, The above method, A method comprising an action of checking from the database evaluation items corresponding to the type of task performed to generate the above result document.

15. In a non-transitory computer-readable recording medium, An action to obtain an original document and a result document generated by performing a task on said original document; An action for determining evaluation items for evaluating the tasks performed to generate the above result document; An action of generating a prompt for requesting an AI model to evaluate the task, the action including at least some of the original document, the result document, and the determined evaluation items; An action of transmitting the generated prompt to the AI model; and A recording medium storing instructions for performing an operation of obtaining a first evaluation result for the performed task from the AI model.

Images